In your opinion, which criteria does a MOOC have to fulfill in order to be considered best practice?

Pierre Dillenbourg: Most discussions about MOOCs are concerned with the videos, namely their quality, their length, etc. Of course we want the best quality, especially audio quality, but students do not join MOOCs because the video is nice. Overall they expect high-impact contents. Actually, videos are only the visible tip of the iceberg.

As best practice, I would qualify  a MOOC with rich activities between the videos: running analyses in a statistical platform, using a simulation, conducting remote sensing experiments, exploring a repository of museum archives, annotating maps of future cities, etc. What students do is what they learn. The best learning activities will also connect what they learn to their workplace.

How can a user recognize a good or bad MOOC as quickly as possible, and what are the criteria?

Pierre Dillenbourg: They are simple cues, such as the name of the university that produced the MOOC or the instructor’s scientific reputation. But potential participants should also look at the opinions of those who have completed the MOOC. Recommender systems for MOOCs are emerging, which will be useful - but of course are exposed to the same biases as hotel ratings.

In your opinion, do the high participant drop-out rates that everyone complains about have anything to do with the quality of MOOCs, or are they due to the participants’ lack of motivation?

Pierre Dillenbourg: The rate of drop out is due first to a misunderstanding of the word ‘participant’. Users need to register for a MOOC to be able to see its contents. More than one third of them watch the first video and then decide that the MOOC does not match their expectations and hence do not engage at all. They do not drop out; they never really started. Then, another third takes the MOOC as a wiki: they don’t aim to get a certificate but select some MOOC components that interest them.

EdX has found that twenty percent of their MOOC participants are teachers. One should only consider those who really want to complete the MOOC. Among them, some will drop out because of low quality, but, the main reason to drop out is probably the workload: our MOOCs are as demanding as any EPFL on-campus class, and two thirds of MOOC participants have a job. Furthermore, many of them probably have a family, and keeping the MOOC tempo is not trivial. This is why both Coursera and EdX now offer courses for which learners can choose their own learning pace.

Can we expect quality standards for MOOCs in the near future?

Pierre Dillenbourg: Yes. Several standards will probably emerge from different categories of MOOCs. For instance, some of our MOOCs aim to better prepare freshmen to succeed in our very selective first undergraduate year (e.g. our introductory MOOC in Physics). Other MOOCs aim to give access to specific, advanced scientific knowledge (e.g. our MOOCs on neuronal dynamics and plasma physics). Finally, some MOOCs add value on the job market (e.g. our MOOC on the programming language SCALA).

So, actually, very different types of learning events fall under the rubric “MOOC”, and their quality has to be assessed by specific criteria. Nonetheless, sooner or later, universities will give European Credit Transfer and Accumulation System (ECTS) credits, and therefore some rules of assessing MOOC quality will be necessary. The day we are able to anchor this quality control into the ECTS system, the Europe of Bologna will lead worldwide education.